Studies in cerebral circulation have suggested that brain metabolism must be measured in association with blood flow in order to understand many pathophysiological entities. Currently, no method exists whereby these can be studied serially for long intervals in the same animal. Our preliminary anatomical and angiographical studies have indicated that certain paired cerebral veins drain only blood from cortex supplied by easily identified cerebral arteries. These veins have been cannulated bilaterally without altering flow, in their territory of distribution. They will stay patent in a heparinized animal for several days, during which time the venous flow rate can be measured. Samples of this venous blood can be analyzed periodically for oxygen, glucose, K ion, H ion, etc. When compared to the arterial concentrations of the substances and the regional blood flow, one can calculate the regional metabolic rate of tissue in this inflow-outflow system (vascular-metabolic unit). This system can be subjected to ischemia, venous obstruction, and pharmacologic alterations in metabolism. The effects of these simulated pathophysiologic entities can be studied. Hopefully, metabolic indices for reversible and irreversible neuron damage can be calculated. Ultimately, techniques which lower brain metabolism will be applied to the system with the hope that irreversible neuron damage can be reduced. The vascular-metabolic unit concept can be developed for the study of other disorders which may affect specific regions of brain (trauma, epilepsy, etc.).